234 



STEPHEN DEM. GAGE 



incubation and at 30 C. after 24 hours' incubation, was very similar, 

 and although the numbers of bacteria determined at 30 C. were 

 smaller than at 20 C., they were fully as significant. The substi- 

 tution of the 24-hour count at 30 for the usual 20 count in labora- 

 tories connected with water filtration plants would result, therefore, 

 in a reduction of the period required to complete a bacterial anal- 

 ysis, without causing any material change in the theories upon which 

 the interpretation of the results are based. The value of counts 

 at 40 C. made at the same time as counts at 20 or 30 should not 

 be underestimated, since the group of bacteria determined at this 

 temperature undoubtedly represent more closely the pathogenic 

 bacterial content of the water than is the case with counts at lower 

 temperatures. It would be inadvisable however, to employ 40 

 counts exclusively in water work at the present time, until we have 

 a broader understanding of their complete significance, although 

 such counts are used exclusively in some laboratories in the control 

 of milk supplies. The average results of counts at the various tem- 

 peratures with eight different waters are shown in the following 

 table: 



TABLE 2. 



AVERAGE NUMBER OF BACTERIA AND ACID-PRODUCERS DEVELOPING AT 20. 30, AND 40 C. WITH 

 DIFFERENT CLASSES OF WATER. 



Bacterial ratios, jo series. In Table 3 are shown the various 

 bacterial ratios for the determinations at 20, 30, and 40 C. Cer- 

 tain distinctions between the different waters are brought out by 

 these ratios which do not appear in the numbers of bacteria. In 

 the first two columns are shown the per cent which the numbers 

 of bacteria determined at 30 and 40 after 24 hours' incubation 

 are of the total bacteria determined at 20 after four days' incuba- 



